1. Field of the Invention
The present invention relates to an optical filter and an optical device provided with this optical filter.
2. Description of the Related Art
In a digital still camera employing an imaging device such as a CCD (hereafter a digital still camera is simply referred to as a "DSC" in this specification), "beat" interference may occur as a result of a certain relationship between the spatial frequency of the subject image and the repetitive pitch of dot-type on-chip color separation filters provided at the front surface of the imaging device. In order to prevent any false color signals from being generated by the beat, i.e., in order to prevent the so-called "color moire," an optical low-pass filter is provided between the taking lens and the imaging device. The optical low-pass filter, which is constituted by employing a birefringent plate achieving birefringence, reduces the generation of the beat through the birefringent effect provided by the birefringent plate. Normally, quartz is employed to constitute the birefringent plate.
Japanese Examined Patent Publication No. 1994-20316 proposes an optical low-pass filter employing two birefringent plates such as that described above, which is suited for application in an imaging device provided with dot-type on-chip color separation filters. This optical low-pass filter is constituted by enclosing a quarter-wave plate between two birefringent plates with the directions in which the image becomes shifted through the birefringence offset by approximately 90.degree. from each other.
Now, the so-called direct image forming system, in which the imaging device is directly provided at the primary image forming plane of the taking lens without employing a reduction lens system or the like is becoming the mainstay in single lens reflex type DSCs that allow interchange of the taking lens among DSCs in recent years. The advent of the direct image forming system has been realized through the utilization of imaging devices having a large image area of approximately 15.5 mm.times.22.8 mm that have been manufactured in recent years to replace 2/3" size (approximately 6.8 mm.times.8.8 mm) and 1" size (approximately 9.3 mm.times.14 mm) imaging devices that have been conventionally used in television cameras and the like. With this size of image area available, an image plane having a size (approximate aspect ratio 2:3=15.6 mm.times.22.3 mm), which is comparable to the image plane size of the C-type silver halide film IX240 system (APS), is achieved. By employing an imaging device achieving a relatively large image area, it becomes possible to adopt a camera system that employs the 135-type photographic film in a DSC. To explain this point, providing a 2/3" size or 1" size imaging device at the field of a camera using the 135-type film only achieves a small image plane size for the imaging device compared to the image plane size of the 135-type film (24 mm.times.36 mm). As a result, a large difference will manifest in the angle of field achieved by a taking lens having a specific focal length, to cause the photographer to feel restricted. This problem becomes eliminated as the image area of the imaging device increases and becomes closer to the image plane size of the 135-type film.
However, as the image area in a single lens reflex type DSC, which forms the primary image with the taking lens at an image device directly, increases, the problems explained below arise to a degree to which they cannot be neglected.
Imaging devices in DSCs in recent years have evolved in two directions, i.e., toward a higher concentration of pixels and toward a larger image plane. When the number of pixels is increased to exceed 1 million pixels while maintaining the size of the image plane at approximately 1/3" to 1/2" as in the prior art, the pixel pitch becomes reduced. For instance, in an imaging device having approximately 1,300,000 pixels, with its image plane size at approximately 1/3", the pixel pitch is approximately 4 .mu.m. Generally speaking, the pixel pitch "p" at an imaging device and the thickness "t" of the birefringent plates constituting the optical low-pass filter which is employed to support the pixel pitch "p" achieve the relationship expressed through the following equation (1) EQU p=t(ne.sup.2 -no.sup.2)/(2ne.times.no) (1)
with
When a quartz plate, which is most commonly employed to constitute a birefringent plate, is used in an imaging device with a pixel pitch of approximately 4 .mu.m, the thickness "t" required of the quartz plate is concluded to be approximately 0.7 mm by working backward with "p" in equation (1) set at 4 .mu.m, since the refractive indices of quartz for light having a wavelength of 589 nm are ne=1.55336 and no=1.54425. Since the thickness of the quarter-wave plate needs to be approximately 0.5 mm regardless of the pixel pitch p, the entire thickness achieved when constituting an optical low-pass filter by pasting together three plates, i.e., two quartz plates (birefringent plates) and one quarter-wave plate, will be approximately 2 mm.
However, when the area of the photosensitive surface of an imaging device increases, as in the case of, in particular, an imaging device employed in a single lens reflex type DSC, it becomes necessary to increase the thickness of the optical low-pass filter for the reasons detailed below.
While the size of the image plane of a 1/3" imaging device is approximately 3.6 mm.times.4.8 mm, let us now consider an imaging device having an image plane size equivalent to that of the C-type (aspect ratio 2:3=16 mm.times.24 mm) in an IX240 system (advanced photo system (APS)) with silver halide film. When pixels are arrayed at a pixel pitch of approximately 4 .mu.m on this imaging device, the total number of pixels for the entire image plane will exceed 20 million by simple calculation, and it is considered that the current technical level is not high enough to realize such a large number of pixels for practical use from the viewpoints of the yield in imaging device production, the scale and processing speed of the image information processing circuit and the like. As a result, it is assumed that it is appropriate to set the number of pixels at approximately two million and several hundreds of thousands in an imaging device having a large image plane equivalent to that of the APS-C type, which sets the pixel pitch at 10 and several .mu.m.
For instance, when an APS-C size imaging device (16 mm.times.24 mm) is prepared at a pixel pitch set to 12 .mu.m, the number of pixels in the imaging device will be approximately 2,670,000. When constituting the birefringent plates of the optical low-pass filter employed in combination with the imaging device having the pixel pitch of 12 .mu.m with quartz, the thickness of a single quartz plate is calculated to be "t"=2.04 mm by incorporating "p"=12 .mu.m in equation (1). By adding the thicknesses of two such quartz plates and a quarter-wave plate (0.5 mm), the thickness of the optical low-pass filter is calculated to be 4.58 mm, which is more than twice as large as the thickness of an optical low-pass filter (thickness: 2 mm) with the pixel pitch set at 4 .mu.m.
In addition, since the spectral sensitivity of an imaging device is different from the spectral sensitivity of the human eye, an IR blocking filter is normally provided to cut off infrared light within the imaging optical path in a DSC employing an imaging device. This IR blocking filter (thickness: approximately 0.8 mm) is also provided pasted to the optical low-pass filter. Thus, the entire thickness of the optical low-pass filter supporting the pixel pitch of 12 .mu.m will go up to 5.38 mm when the thickness of the IR blocking filter is included.
It is difficult to place an optical low-pass filter having such a thickness between a taking lens and an imaging device. Even in the case of a regular lens shutter type DSC, which does not require any member to be provided between the rear end of the taking lens and the photosensitive surface of the imaging device except for the optical low-pass filter, it must be ensured in design that the minimum value (the so-called back focal distance) of the distance between the rearmost end of the taking lens and the imaging device is larger than the thickness of the optical low-pass filter. Setting the length of the back focal distance of the taking lens larger than the focal length of the taking lens imposes restrictions in terms of the optical design.
Furthermore, in the case of a single lens reflex type DSC, which directly forms an image of the subject achieved by a taking lens on a large size imaging device without employing a reduction lens system, a quick return mirror for switching the optical path between the viewfinder and the imaging system or a fixed semitransparent mirror (beam splitter) is needed between the taking lens and the imaging device. In addition, a mechanical shutter is required for defining an exposure time and for blocking the imaging device from exposure during an image signal read operation at the imaging device. While this structure having a mirror and a shutter provided between the taking lens and its image forming plane is also adopted in a single lens reflex camera that employs regular silver halide film, it is difficult to provide an optical low-pass filter having a thickness exceeding 5 mm in addition while ensuring that it does not present any obstacle in the operation or the mirror at the shutter. It merits particular note that more and more cameras in recent years adopt the autofocus (AF) function, and that a single lens reflex camera with the AF function adopts a structure having a sub mirror provided to the rear of the quick return mirror, i.e., between the quick return mirror and the shutter, to guide light flux to a focal point detection device. This makes it even more difficult to position an optical low-pass filter having a thickness exceeding 5 mm.
In addition to the problem of an increased thickness of the optical low-pass filter resulting from a larger pixel pitch in a larger imaging device as described above, another problem arises as detailed below.
Normally, the length of the air equivalent optical path achieved when light is transmitted and advances through a medium having a thickness "t" and a refractive index "n" is expressed as t/n. In other words, the air equivalent optical path lengths achieved when light advances through media having the same refractive index "n" but having different thicknesses "t", vary. Now, let us consider light emitted from one point on the optical axis of a photographic optical system toward an imaging device to reach the center of the image plane of the imaging device and light emitted from the same point on the optical axis of the photographic optical system toward the imaging device to reach the periphery of the image plane.
Since the light that reaches the center of the image plane enters the light entry surface of the optical low-pass filter at almost a right angle, "t" roughly equals the thickness of the optical low-pass filter. In contrast, since the light reaching the periphery of the image plane advances diagonally through the optical low-pass filter, "t" here is larger than the "t" encountered by the light reaching the center of the image plane. Since the lengths of the air equivalent optical paths achieved by light being transmitted through the optical low-pass filter are different for the light reaching the center of the image plane and the light reaching the periphery of the image plane as explained above, a focus misalignment occurs in the direction of the optical axis between the image plane center and the image plane periphery. The degree of this focus misalignment increases as the thickness of the optical low-pass filter increases as described above, which may result in a reduced image quality at the peripheral area of the image plane.
As the size of the imaging device is increased, a problem of foreign matter becoming transferred as explained next, i.e., a problem of foreign matter such as dust and lint adhering to the photosensitive surface of the imaging device to cast a shadow onto the image captured by the imaging device, tends to occur readily, in addition to the problems discussed above. In particular, in an interchangeable lens type DSC in which foreign matter such as dust and lint readily enters the mirror box when the taking lens is detached, this problem is more pronounced.
A similar problem occurs in optical devices such as facsimile machines and image scanners when foreign matter such as dust and lint materialize as a document is transmitted or the document read unit moves, which may become adhered to the vicinity of the photosensitive surface of the photoelectric conversion element or the glass (platen glass) upon which the document is placed to result in a shadow being cast on the input image, as in the interchangeable lens type DSC.
Now, since the crystal of quartz employed to constitute birefringent plates imparts a piezoelectric effect, the crystal itself is caused to become electrically charged readily by vibration or the like. The quartz crystal also has a property that does not allow a stored electrical charge to be discharged easily. In addition, since an insulating material such as plastic, ceramic or the like is employed to constitute the imaging device package, the electrical charge stored at the imaging device cannot be released with ease.
Vibration and air currents occurring as a result of an operation of an optical device sometimes cause the foreign matter discussed above to become suspended inside the optical device, which may ultimately become adhered to the electrically charged birefringent plates, imaging device or the like, as explained above. Consequently, the operator of the optical device is required to clean the optical device frequently to prevent shadows from being cast as explained earlier.